Schema Validator

Structured data is the difference between a page Google ignores and a page that gets a rich result, an AI answer engine citation, and a shopping snippet with a live price. But most schema markup on the web is broken — and the free validators everyone uses only catch a fraction of the real problems. The ToolsVale Schema Validator is the free schema validator built for how search actually works in 2026: it checks JSON-LD, Microdata, and RDFa against the current schema.org vocabulary (v30, released March 25, 2026), verifies eligibility under Google’s 2026 rules, cross-checks every value against what’s visible on your page, and — critically — fetches your URL as five different crawlers including GPTBot, ClaudeBot, and PerplexityBot to show you exactly what AI answer engines see.

This page explains what a schema validator does, what most validators miss, and how our tool catches the schema problems that actually matter for rankings and rich results in 2026. If you’re impatient, scroll up and paste a URL — the validator runs in about 10 seconds and shows results as they arrive.

What is a schema validator?

A schema validator is a tool that checks your website’s structured data markup for two things: whether it conforms to the schema.org vocabulary, and whether it qualifies for search-engine rich results. Structured data itself is machine-readable metadata you embed in your HTML — usually as JSON-LD inside a <script type="application/ld+json"> tag — that tells Google, Bing, ChatGPT, Claude, and other consumers what your page is about in a structured way.

When your Product schema declares a price, a rating, a stock status, and a brand, Google can render your listing in search results as a rich card with stars, a price, and an “In Stock” badge. When your Article schema declares an author with credentials and a publisher with a logo, AI answer engines like ChatGPT and Perplexity can cite you as a source with attribution. When your LocalBusiness schema has valid geo coordinates and opening hours, your storefront can appear in the Google Maps knowledge panel with directions and hours built in.

But structured data only works if it’s correct. A missing required field, a deprecated type, a currency string Google can’t parse, a rating count below the eligibility threshold, or — most commonly in 2026 — a schema value that contradicts what’s on the visible page will silently disqualify you. That’s what a schema validator catches, and it’s why every serious SEO workflow includes a validator step before publishing.

Why most schema validators fall short in 2026

Search evolved fast between 2024 and 2026, and most free schema validators haven’t kept up. Here’s what changed and what most tools still miss:

Content parity became the #1 spam trigger. Google’s March 2026 core update elevated content parity — the requirement that structured data values match what’s visible on the page — from a best practice to a hard ranking signal. If your Product schema says $99 but the visible price is $79 (because you launched a sale but the schema wasn’t updated), Google treats this as structured-data spam and can suppress your rich result entirely. Most validators, including Google’s own Rich Results Test, don’t check for this.

AI answer engines became a real referral source. ChatGPT, Claude, Perplexity, and Gemini increasingly cite structured data to build their answers. Their crawlers — GPTBot, ClaudeBot, and PerplexityBot — often see a different version of your page than Chrome or Googlebot. If you inject schema client-side via Google Tag Manager, React useEffect, or hydration, AI crawlers see zero schema. Most validators fetch as a single user agent and can’t reveal this gap.

Google deprecated seven rich result types in January 2026. BookAction, Course, EmployerAggregateRating, EstimatedSalary, the JobPosting Hiring extension, LearningResource, and Occupation are no longer eligible for rich results. Sites still using them silently lose visibility. Validators that haven’t updated their rules quietly pass these as valid.

FAQPage rich results were removed for most sites in May 2026. Google now only shows FAQPage rich results for well-known government and health authority sites. Every other site with FAQPage schema is spending crawl budget on markup that will never render. Most validators still recommend FAQPage as best practice.

Schema.org released v30 on March 25, 2026. New types, deprecated properties, and refined domains. Validators using older versions (some public tools are still on 2020-2022 vocab) pass or fail the wrong things.

The ToolsVale Schema Validator was built with all five of these shifts as first-class concerns. That’s why it exists.

What this schema validator checks that others don’t

Every free schema validator checks basic syntax — that your JSON-LD parses, that required fields are present. That’s the floor. Here’s what our tool adds on top:

1. Content parity engine

Every value in your structured data — prices, availability, ratings, review counts, dates, addresses, names — is cross-referenced against the visible DOM using RapidFuzz semantic matching. If your schema declares "price": "99.00" but the page displays $79 in the visible price element, we flag PARITY-002 (critical) with both values shown side by side. Same for availability contradictions (“InStock” in schema, “Out of Stock” on page), rating mismatches, and stale dates. Match thresholds are 85% for strong equivalence, 60% for weak — you can trust the flags aren’t false positives from formatting differences.

2. Five-crawler visibility check

Every URL you submit is fetched as five different user agents: Chrome desktop (real users), Googlebot desktop (search), GPTBot (ChatGPT), ClaudeBot (Claude), and PerplexityBot (Perplexity). We compare the schema block count, the entity types extracted, and key values across all five. If any crawler sees materially different content, we flag it as cloaking with the specific differences highlighted. This is the fastest way to catch the “my schema works in Chrome but AI answer engines can’t see it” problem — the biggest single reason products don’t appear in ChatGPT or Perplexity answers today.

3. Schema.org v30 + Google 2026 rule tracking

Our vocabulary database is the March 25, 2026 schema.org release. Our eligibility rules are current with Google’s Search Central documentation as of this month. Deprecated types raise GOOGLE-004. Removed rich result types (FAQPage on non-government sites) raise GOOGLE-005 with the deprecation date and Google’s official reasoning. When Google announces changes, we update the same day.

4. Three-tier fix classification

Every fix we generate is labeled by safety tier so you know what’s safe to apply automatically and what needs human judgment:

• DETERMINISTIC — copy-paste ready fix. Applying it will not break anything. Example: adding a missing required @id field with a computed value based on the page URL.
• SAFE_WITH_REVIEW — the fix is likely correct but needs a quick human check. Example: normalizing a currency string from “US$” to “USD” per ISO 4217.
• SEMANTIC_HUMAN_REQUIRED — the fix needs semantic judgment and can’t be automated safely. Example: writing a meaningful product description to replace a placeholder.

This prevents the automated “fixes” that make things worse — a common failure mode with AI-generated schema tools that ship copy-paste fixes without safety context.

5. Framework and CMS attribution

For every schema block on your page, we identify which system produced it. Our signatures database covers WordPress (with Yoast, Rank Math, All in One SEO, Schema Pro, WP Recipe Maker, The Events Calendar, and WooCommerce plugin detection), Shopify, Webflow, Wix, Squarespace, Drupal, Joomla, Framer, and Ghost. On the JavaScript-framework side we detect Next.js, Nuxt, Gatsby, Astro, SvelteKit, React, Vue, Angular, and Google Tag Manager injection. When you have a bug, you need to know which system produced it so you know which template, plugin, or component to fix. Standard validators never tell you this.

Understanding the three scores

Every URL you validate gets three scores on a 0-100 scale, each with a breakdown:

Schema Health measures vocabulary correctness against schema.org v30. Are the types real? Are required properties present? Are property values in the correct format (URLs where URLs are expected, ISO 8601 for dates, ISO 4217 for currency codes)? Are there any orphan nodes in your @graph, cycles, or duplicate @ids? This score is deterministic — the same page produces the same score every time.

Google Eligibility measures how well your schema qualifies for Google rich results under current 2026 rules. Does your Product have enough reviews (Google’s threshold is 3+ reviews for AggregateRating rich results)? Is your Article’s dateModified present and valid? Does your Recipe include the required cookTime, prepTime, and nutrition fields? Deprecated types cost points here. So does using types Google never supported.

AI Readiness measures how discoverable your schema is by AI answer engines specifically. This is a simplified 5-criteria model (we deliberately cut it from a proposed 12 criteria to prevent noise): (1) Organization completeness — canonical Organization entity with logo, sameAs, contactPoint; (2) SameAs reachability — do linked social profiles actually resolve to 200 OK, cached 24 hours; (3) Author entity depth — real Person entity with credentials, not just a name string; (4) Content parity; (5) Server-side rendering — schema present in raw HTML, not injected by JavaScript. AI answer engines weight these signals heavily when deciding what to cite.

Content parity: the March 2026 game-changer

Of everything the ToolsVale Schema Validator does, the content parity check is the most important — because it catches the single class of problem that Google now punishes hardest and that no other free validator catches.

Here’s the mechanism. When Google’s crawler fetches your product page, it does two things in parallel: it parses your structured data (getting a schema price like $99, an availability status like InStock, a rating of 4.5 with 127 reviews), and it parses the visible DOM (getting the price element showing $79, an “Out of Stock” badge, and no visible review count). If those two views disagree — if the schema is claiming something the visible page contradicts — Google’s March 2026 core update treats it as intentional spam. The rich result gets suppressed, and in severe cases the whole URL’s ranking gets a soft demotion.

Most parity mismatches aren’t intentional. They come from three common patterns: (1) sale prices that update the visible page but not the schema (or vice versa), (2) inventory status that changes in the product feed but not in the ecommerce template, and (3) review counts that get cached in the schema while the actual review widget shows a fresher number. Every one of these is fixable in minutes once you know it’s there — but until now, you had no way to know it’s there.

Our Content Parity Engine solves this by extracting every semantic value from your structured data and searching the visible DOM for it using RapidFuzz fuzzy matching. Exact matches pass silently. Values not found trigger a critical PARITY-002 with both the schema value and the visible value shown, plus a fix suggestion (usually “update your schema to match the visible page” or “update your visible template to match the intended schema”). We tuned the fuzzy thresholds (85% for strong equivalence, 60% for weak) so cosmetic differences like “$99.00” vs “$99” pass without flagging, but real contradictions like “$99” vs “$79” fire immediately.

AI crawler visibility: schema for the ChatGPT era

AI answer engines don’t render JavaScript. When ChatGPT wants to answer “what’s the best home espresso machine under $500,” GPTBot fetches candidate pages, extracts their structured data, and cites the results — but only if that structured data is present in the raw HTML response. Any schema you inject client-side, via Google Tag Manager or React hydration or a WordPress plugin that runs on DOMContentLoaded, is invisible to AI crawlers.

This is the single biggest reason products, articles, and reviews don’t appear in AI answer engine citations today. And it’s silent — nothing in your Google Search Console flags it, nothing in your analytics shows the missed opportunity, and Chrome and Googlebot see the schema fine (Googlebot renders JavaScript before extracting structured data, so it works there). You only find out when you notice your competitor gets cited in ChatGPT for the query you should be winning.

Our five-crawler check reveals this in one fetch. We hit your URL as Chrome desktop, Googlebot desktop, GPTBot, ClaudeBot, and PerplexityBot in parallel, extract structured data from each response, and compare. If Chrome sees 8 schema blocks and GPTBot sees 0, you know the answer immediately: your schema is client-side only, and you need to move it to server-side rendering. We flag this as SOURCE-003 (high JS injection risk) with the specific framework or plugin identified and a recommended fix path.

For any site targeting AI-answer-engine traffic in 2026 — and that’s now essentially every site — this check should be part of every content publish workflow.

Schema.org v30 and Google 2026 rules

Structured data rules change quarterly. Between January 2026 and August 2026, Google made two major changes that affect nearly every ecommerce and content site:

January 2026 deprecations. Seven rich result types were deprecated: BookAction, Course, EmployerAggregateRating, EstimatedSalary, the JobPosting Hiring organization extension, LearningResource, and Occupation. Sites still using these are wasting crawl budget on markup that will never render — the schema still parses, but the rich result won’t show. Our validator flags these as GOOGLE-004 with the deprecation date and a recommendation to remove or downgrade the type.

May 2026 FAQPage removal. Google restricted FAQPage rich results to well-known government agencies and recognized health authorities. Every other site with FAQPage schema is running dead markup. If your site is a commerce site, blog, SaaS product page, or agency site with FAQPage schema, we flag GOOGLE-005 and recommend removing the FAQPage wrapper (you can keep the FAQ content itself — just drop the schema).

Schema.org released v30 on March 25, 2026 with new types, refined property domains, and clarified value expectations. Our vocabulary database mirrors that release. When schema.org publishes v31, we’ll update within a week of release.

How to use the schema validator

Scroll up to the tool card, paste one to twenty-five URLs, and click Run validator. Results stream in as they arrive — you’ll see each URL’s card appear with three score gauges, then the issue list expands below with categorized findings.

The Issues section is open by default and sorted by severity (critical first). Each issue shows a stable rule ID (like PARITY-002 or GOOGLE-004), an evidence snippet showing the exact offending markup, a confidence score, and — where applicable — a copy-paste fix in a dark code block. Click Details & fix on any issue to see the mechanism explanation and the safety tier of the suggested fix.

Below Issues, you can expand Extraction (block counts by format, entity types seen, syntax errors), Source attribution (CMS, plugin, framework detection with JS injection risk), Crawler visibility (per-UA status codes and schema block counts), Schema graph (nodes, edges, orphans), and AI Readiness detail (5-criteria breakdown with top recommendations).

Every audit can be shared via a 48-hour link — click Create shareable link in the Share section. Anyone with the URL can view the full report without signing up. Useful for handing off to developers, clients, or SEO auditors.

Common schema mistakes we catch

After validating hundreds of production sites, these are the mistakes we see most often — and that most other validators miss:

• Sale prices in the visible page but not in the schema. Extremely common on WooCommerce and Shopify. Our PARITY-002 catches this immediately.
• Currency codes as symbols instead of ISO 4217. “$” or “US$” instead of “USD”. Google requires ISO 4217. VOCAB-008 flags this with the corrected value.
• Review counts below Google’s threshold. AggregateRating with only 1-2 reviews doesn’t qualify for rich results. GOOGLE-004 tells you the minimum (3) and recommends removing the AggregateRating until you meet it.
• priceValidUntil dates in the past. Products with expired price validity get suppressed from rich results. GOOGLE-004 catches this.
• Missing dateModified on Article. Google requires this for freshness signals. ARTICLE-002 flags it.
• Product without brand or GTIN. PRODUCT-001 catches missing GTIN, PRODUCT-002 catches missing brand — both required for shopping rich results.
• LocalBusiness with swapped lat/lng. Common when copying coordinates from Google Maps in the wrong order. LOCAL-002 detects when latitude and longitude values look swapped based on ranges.
• Article headline over 110 characters. Google truncates at 110 for AMP and card display. ARTICLE-001 flags with the character count.
• Recipe without cookTime or nutrition. Both required for Recipe rich results as of the 2024 update. RECIPE-001 and RECIPE-002 flag these.
• Duplicate @id across your @graph. Causes downstream graph resolution failures. GRAPH-003 catches this with both offending entities shown.
• Orphan nodes (defined but never referenced). GRAPH-005 lists them so you can remove or wire them up.
• sameAs URLs that 404 or 301 to unrelated domains. SAMEAS-001 checks reachability with 24-hour cache; SAMEAS-002 flags redirect chains.

Who this schema validator is for

The ToolsVale Schema Validator is built for SEO teams, ecommerce operators, content publishers, and developers who need to ship correct structured data at scale. If you’re auditing a single page for a client, batch-checking an entire product category before a launch, or building schema into a CI/CD pipeline, this tool is designed for you. If you’re new to structured data and just want to add JSON-LD to your first blog post, you’ll probably start with a generator like the built-in features of Rank Math or Yoast — and then come here to verify what you generated actually works.

It’s also built for anyone whose SEO strategy now includes AI answer engines. If you’re wondering why your competitor gets cited in ChatGPT for queries you should be winning, run their URL and yours through the validator. Nine times out of ten, the answer is in the AI Readiness score.

Get started

Scroll to the tool card above, paste a URL, and hit Run validator. The first audit takes about 10 seconds. The tool is free forever, no signup, no watermark, no upgrade prompts. For related audits, see our Meta Tag Checker, Open Graph Checker, Broken Link Checker, and Google Index Checker — together they cover the full pre-publish SEO checklist.